Bringing clarity and transparency to the consultative process underpinning the implementation of an ethics framework for AI-based healthcare applications: a qualitative study

Abstract

Artificial intelligence (AI) has been applied in healthcare to address various aspects of the COVID-19 crisis including early detection, diagnosis and treatment, and population monitoring. Despite the urgency to develop AI solutions for COVID-19 problems, considering the ethical implications of those solutions remains critical. Implementing ethics frameworks in AI-based healthcare applications is a wicked issue that calls for an inclusive, and transparent participatory process. In this qualitative study, we set up a participatory process to explore assumptions and expectations about ethical issues associated with development of a COVID-19 monitoring AI-based app from a diverse group of stakeholders including patients, physicians, and technology developers. We also sought to understand the influence the consultative process had on the participants’ understanding of the issues. Eighteen participants were presented with a fictitious AI-based app whose features included individual self-monitoring of potential infection, physicians’ remote monitoring of symptoms for patients diagnosed with COVID-19 and tracking of infection clusters by health agencies. We found that implementing an ethics framework is systemic by nature, and that ethics principles and stakeholders need to be considered in relation to one another. We also found that the AI app introduced a novel channel for knowledge between the stakeholders. Mapping the flow of knowledge has the potential to illuminate ethical issues in a holistic way.